Developing first-class software for first-class businesses.

"Anand and others from GuardX have been delivering big data strategy and implementation for us for more than 7 years – always a cut above the rest."

- ED, Retail Risk Reporting
Tier 1 US Bank

"Align your technology strategy and requirements with the performance of your business by understanding people, processes and technology in your security program."

– MD, Regulatory Risk Technology
Tier 1 Global Bank

"Information extraction system for our semi structured reports was exemplary and easy to use."

- EVP
Major Oil drilling corporation

Unstructured Data Analysis Project

  • Goal: extract information from Oil Well exploration documents which were semi structured.
  • Analyzed 100,000+ documents to train standard ML models like DF-IDF, and our proprietary models.
  • Model adjustments were made on the fly with incoming documents via feedback loop.
  • Information storage & visualization made it a very practical system in use, currently in production.

Twitter Project

  • Goal: needed to guard against book sweeps by prop shops on events reported on Twitter before it hits market.
  • Effectively created a large twitter feed data (~50TB / year) Union with our options tick plant (~200TB compressed / year), managed by a staff of 4 technologist for entire US options universe (~1500 names).
  • Created Domain Specific Language to integrate complex rule into the system for signal strength measure.
  • Large Training set based learning to extract signals on the fly, correlated historically to market movements.

Retail Risk Project

  • Goal: Customer account data storage from 30+ countries.
  • Domain Language based modeling for retail risk reporting.
  • Data validation and enrichment using DSL.
  • Traceable, auditable high performance model workbench built on Hadoop / Spark.
  • Micro Services for SAS / R interaction with Hadoop.